intl: Support the AIX 7 locale names.

* gettext-runtime/intl/explodename.c (_nl_find_language): Remove function.
(_nl_explode_name): Inline it here. On AIX, lowercase the language and map a
script identifier to a modifier.
* NEWS: Mention it.

diff --git a/NEWS b/NEWS
index f2764d4eef1958ed9f5b2f8428252958f60a4ab8..cdbb16746c23555e70bb1e16917f5c349ce92d9e 100644 (file)
--- a/NEWS
+++ b/NEWS
@@ -1,5 +1,11 @@
 Version 0.21.1 - October 2022
 
+* Runtime behaviour:
+  - On AIX, locale names with a script or with an uppercase language are now
+    supported.
+    For example, sr_Cyrl_RS.UTF-8 is treated like sr_RS.UTF-8@cyrillic, and
+    EN_US.UTF-8 is treated like en_US.UTF-8.
+
 * The base Unicode standard is now updated to 14.0.0.
 
 * Portability:

diff --git a/gettext-runtime/intl/explodename.c b/gettext-runtime/intl/explodename.c
index 57c3e4f2d9d5a442c4aa5ee1b9a83289888b7c5b..661266ea30e3b54ae89ea7a1eeee178664aeb82a 100644 (file)
--- a/gettext-runtime/intl/explodename.c
+++ b/gettext-runtime/intl/explodename.c
@@ -1,4 +1,4 @@
-/* Copyright (C) 1995-2016 Free Software Foundation, Inc.
+/* Copyright (C) 1995-2016, 2020 Free Software Foundation, Inc.
    Contributed by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1995.
 
    This program is free software: you can redistribute it and/or modify
@@ -35,20 +35,6 @@
 
 /* @@ end of prolog @@ */
 
-/* Split a locale name NAME into a leading language part and all the
-   rest.  Return a pointer to the first character after the language,
-   i.e. to the first byte of the rest.  */
-static char *_nl_find_language (const char *name);
-
-static char *
-_nl_find_language (const char *name)
-{
-  while (name[0] != '\0' && name[0] != '_' && name[0] != '@' && name[0] != '.')
-    ++name;
-
-  return (char *) name;
-}
-
 
 int
 _nl_explode_name (char *name,
@@ -64,23 +50,84 @@ _nl_explode_name (char *name,
   *codeset = NULL;
   *normalized_codeset = NULL;
 
-  /* Now we determine the single parts of the locale name.  First
-     look for the language.  Termination symbols are `_', '.', and `@'.  */
+  /* Determine the individual parts of the locale name.
+     Accept the XPG syntax
+
+             language[_territory][.codeset][@modifier]
+
+     On AIX systems, also accept the same syntax with an uppercased language,
+     and a syntax similar to RFC 5646:
+
+             language[_script]_territory[.codeset]
+
+     where script is a four-letter code for a script, per ISO 15924.
+   */
+
   mask = 0;
-  *language = cp = name;
-  cp = _nl_find_language (*language);
 
-  if (*language == cp)
+  /* First look for the language.  Termination symbols are `_', '.', and `@'.  */
+  *language = name;
+
+  cp = name;
+  while (cp[0] != '\0' && cp[0] != '_' && cp[0] != '@' && cp[0] != '.')
+    ++cp;
+
+  if (cp == name)
     /* This does not make sense: language has to be specified.  Use
        this entry as it is without exploding.  Perhaps it is an alias.  */
-    cp = strchr (*language, '\0');
+    cp = strchr (name, '\0');
   else
     {
       if (cp[0] == '_')
        {
+         *cp++ = '\0';
+#if defined _AIX
+         /* Lowercase the language.  */
+         {
+           char *lcp;
+
+           for (lcp = name; lcp < cp; lcp++)
+             if (*lcp >= 'A' && *lcp <= 'Z')
+               *lcp += 'a' - 'A';
+         }
+
+         /* Next is the script or the territory.  It depends on whether
+            there is another '_'.  */
+         char *next = cp;
+
+         while (cp[0] != '\0' && cp[0] != '_' && cp[0] != '@' && cp[0] != '.')
+           ++cp;
+
+         if (cp[0] == '_')
+           {
+             *cp++ = '\0';
+
+             /* Next is the script.  Translate the script to a modifier.
+                We don't need to support all of ISO 15924 here, only those
+                scripts that actually occur:
+                  Latn -> latin
+                  Cyrl -> cyrillic
+                  Guru -> gurmukhi
+                  Hans -> (omitted, redundant with the territory CN or SG)
+                  Hant -> (omitted, redundant with the territory TW or HK)  */
+             if (strcmp (next, "Latn") == 0)
+               *modifier = "latin";
+             else if (strcmp (next, "Cyrl") == 0)
+               *modifier = "cyrillic";
+             else if (strcmp (next, "Guru") == 0)
+               *modifier = "gurmukhi";
+             else if (!(strcmp (next, "Hans") == 0
+                        || strcmp (next, "Hant") == 0))
+               *modifier = next;
+             if (*modifier != NULL && (*modifier)[0] != '\0')
+               mask |= XPG_MODIFIER;
+           }
+         else
+           cp = next;
+#endif
+
          /* Next is the territory.  */
-         cp[0] = '\0';
-         *territory = ++cp;
+         *territory = cp;
 
          while (cp[0] != '\0' && cp[0] != '.' && cp[0] != '@')
            ++cp;
@@ -91,8 +138,8 @@ _nl_explode_name (char *name,
       if (cp[0] == '.')
        {
          /* Next is the codeset.  */
-         cp[0] = '\0';
-         *codeset = ++cp;
+         *cp++ = '\0';
+         *codeset = cp;
 
          while (cp[0] != '\0' && cp[0] != '@')
            ++cp;
@@ -111,16 +158,16 @@ _nl_explode_name (char *name,
                mask |= XPG_NORM_CODESET;
            }
        }
-    }
 
-  if (cp[0] == '@')
-    {
-      /* Next is the modifier.  */
-      cp[0] = '\0';
-      *modifier = ++cp;
+      if (cp[0] == '@')
+       {
+         /* Next is the modifier.  */
+         *cp++ = '\0';
+         *modifier = cp;
 
-      if (cp[0] != '\0')
-       mask |= XPG_MODIFIER;
+         if (cp[0] != '\0')
+           mask |= XPG_MODIFIER;
+       }
     }
 
   if (*territory != NULL && (*territory)[0] == '\0')