Name: ###@###.### Date: 08/19/96


Java compiler issues the incorrect error message about Numeric underflow for
the legal floating-point literal:

----------ref:compile.java(4/211)----------
/net/novo33/export/ld32/spn/java_tests/tests/lang/LEX/lex035/lex03502/lex03502.java:13: Numeric underflow.
  if(1.1e1!=0000000000000000000000000000000000000000000000000000000000000001.1E1)
            ^
1 error


Test:
-----

package javasoft.sqe.tests.lang.lex035.lex03502;
import java.io.PrintStream;

class lex03502 {
public static void main (String args []) {
System.exit(run(args,System.out));
}

public static int run(String args[],PrintStream out) {
  if(1.1e1!=0000000000000000000000000000000000000000000000000000000000000001.1E1)
  {
System.out.println("1.1e1!=0000000000000000000000000000000000000000000000000000000000000001.1E1");
  return 2;
  }
  else
  {
System.out.println("1.1e1==0000000000000000000000000000000000000000000000000000000000000001.1E1");
  return 0;
  }
 }
}


C and C++ compilers don't provide the analogous error message as Java compiler
and consider that the values of these floating-point literals are equal:
 
#include <stdio.h>
int main () {
float i;
  if(1.1e1!=0000000000000000000000000000000000000000000000000000000000000001.1E1)
  printf("1.1e1!=0000000000000000000000000000000000000000000000000000000000000001.1E1");
else printf("1.1e1==0000000000000000000000000000000000000000000000000000000000000001.1E1");
}


novo31% cc lex03502.c
novo31% a.out
1.1e1==0000000000000000000000000000000000000000000000000000000000000001.1E1
novo31% CC lex03502.c
novo31% a.out
1.1e1==0000000000000000000000000000000000000000000000000000000000000001.1E1
======================================================================