Fix BZ 69521 - allow Korean (and other) characters in EL identifiers

Regenerate the parser from the JJTree file to pick up the improved
definition of Identifier. It is now identical to Java Language
Identifier except for also allowing identifiers to start with '#'.
Note that while the parser allows identifiers to start with '#', such
identifiers will still be rejected during the validation process.

https://bz.apache.org/bugzilla/show_bug.cgi?id=69521

java/org/apache/el/parser/ELParserConstants.java

@@ -118,9 +118,9 @@ public interface ELParserConstants {
     /** RegularExpression Id. */
     int IMPL_OBJ_START = 58;
     /** RegularExpression Id. */
-    int LETTER = 59;
+    int JAVALETTER = 59;
     /** RegularExpression Id. */
-    int DIGIT = 60;
+    int JAVADIGIT = 60;
     /** RegularExpression Id. */
     int ILLEGAL_CHARACTER = 61;
 
@@ -138,7 +138,7 @@ public interface ELParserConstants {
             "\":\"", "\";\"", "\",\"", "\">\"", "\"gt\"", "\"<\"", "\"lt\"", "\">=\"", "\"ge\"", "\"<=\"", "\"le\"",
             "\"==\"", "\"eq\"", "\"!=\"", "\"ne\"", "\"!\"", "\"not\"", "\"&&\"", "\"and\"", "\"||\"", "\"or\"",
             "\"empty\"", "\"instanceof\"", "\"*\"", "\"+\"", "\"-\"", "\"?\"", "\"/\"", "\"div\"", "\"%\"", "\"mod\"",
-            "\"+=\"", "\"=\"", "\"->\"", "<IDENTIFIER>", "<FUNCTIONSUFFIX>", "\"#\"", "<LETTER>", "<DIGIT>",
+            "\"+=\"", "\"=\"", "\"->\"", "<IDENTIFIER>", "<FUNCTIONSUFFIX>", "\"#\"", "<JAVALETTER>", "<JAVADIGIT>",
             "<ILLEGAL_CHARACTER>", };
 
 }