Building an interpreter for a small programming language using JavaCC

I built a lexer and a parser, using JavaCC, for a small programming language. An example of a valid program in the programming language is:

DEF MAIN { FUNCTION1(4) };
DEF FUNCTION1 x { x+3 } ;

In the above example I want the program to output 7 but to do this a call to FUNCTION1 needs to be made. I know that I have to store intermediate representations somewhere in order to achieve this. How can I do that with Hash Maps?

Suppose the intermediate representation for a function is represented by an object of class Function. The you could have a symbol table that maps function names (presumably represented by objects of type String) to Function objects. The symbol table can be represented by a hash map, e.g.

class SymbolTable extends HashMap<String, Function> { }

You can create an empty symbol table at the start and then fill it as you parse. E.g., something like this.

SymbolTable Program(  )
{
    SymbolTable table = new SymbolTable() ;
}
{
    ( FunctionDefinition( table ) )*
    <EOF>
    { return SymbolTable ; }
}

void FunctionDefinition( SymbolTable table )
{
    Function f ;
    String name ;
}
{
    <DEF>
    name = <NAME>
    f = FunctionBody()
    ";"
    {
      if( table.hasKey( name ) { ...do something... } ;
      else { table.add(name, f) ; }
    }
}