Name: ccC48265 Date: 01/10/98


When compiling and running the following test case that uses
java.io.InputStream.skip()
---------------------------------------------
import java.io.*;

class Y extends java.io.InputStream {
  public int read() {
    return 'a';
  }

  public static void main(String[] args) throws java.io.IOException {
    Y x = new Y();
    long i = 0x100002222L;
    System.out.println("i = 0x" + Long.toHexString(i));
    System.out.println("i = " + i);
    i = x.skip(i);
    System.out.println("i = 0x" + Long.toHexString(i));
    System.out.println("i = " + i);
  }
}
-------------------------------------------
the outcome is:

i = 0x100002222
i = 4294976034
i = 0x2222
i = 8738

It appears that java.io.InputStream.skip argment(n) ignores the
upper 32bit part of (n) because of "new byte[(int)]". However,
section 2.13.10 on "The Java(TM) Application Programming Interface
vol.1" mentions that "The _skip_ method of _InputStream_ creates
a byte array of length _n_ ...".
=======================================================
###@###.### (Jan 10, 1998):
See also bug 4040408.
(Review ID: 23106)
======================================================================