LeDataStrteam consists of Little-endian replacements for DataInputStream,
DataOutputStream and RandomAccessFile you can use in your own Java
programs.  They cannot be run standalone.  They work just like
DataInputStream, DataOutputStream and RandomAccessFile except they work
with little-endian binary data.  Normally Java binary I/O is done with
big-endian data, with the most significant byte of an integer or float
first.  Intel and Windows 95 tend to work with little endian data in
native files.  LEDataInputStream, LEDataOutputStream and
LERandomAccessFile will let you read and write such files.  Source code
provided.
I repeat, LeDataStream is code you insert into your own Java programs.  It
is not a stand-alone utility.  You need basic Java programming skills to
use it.  It lives in package com.mindprod.ledatastream so its *.java files
would live in com\mindprod\ledatastream in the same source code tree as
your own code.
You would treat the LeDataStream *.java files just as if you had written
the code yourself.
e.g. import com.mindprod.ledatastream.LEDataInputStream;
The java.nio (new I/O) package that is new with JDK1.4 has ByteBuffer and
friends that support big and little endian.  That way you don't need
LEDataStream.  LEDataStream is considerably simpler, howewer.
Note that LEDatastream.read(byte b[], int off, int len) is exactly the
same as DataStream.read((byte b[], int off, int len).  With bytes, there
is no difference between big and little endian data.  LEDatastream.read
does NOT reverse the order of the bytes read!
For background see http://mindprod.com/jgloss/endian.html
Why the egg icon?  In Jonathan Swift's Gulliver's Travels, two countries
fought over which end of the end was best to break it, the little or big
end, mirroring the struggle between the users of little and big endian
binary formats.  See http://mindprod.com/jgloss/endian.html