Unsafe Mobile Code: Inner Class

The program violates secure coding principles for mobile code by making use of an inner class.

Inner classes quietly introduce several security concerns because of the way they are translated into Java bytecode. In Java source code, it appears that an inner class can be declared to be accessible only by the enclosing class, but Java bytecode has no concept of an inner class, so the compiler must transform an inner class declaration into a peer class with package level access to the original outer class. More insidiously, since an inner class can access private fields in their enclosing class, once an inner class becomes a peer class in bytecode, the compiler converts private fields accessed by the inner class into protected fields.

Example 1: The following Java Applet code mistakenly makes use of an inner class.

public final class urlTool extends Applet {
private final class urlHelper {
	...
}
	...
}

Mobile code, in this case a Java Applet, is code that is transmitted across a network and executed on a remote machine. Because mobile code developers have little if any control of the environment in which their code will execute, special security concerns become relevant. One of the biggest environmental threats results from the risk that the mobile code will run side-by-side with other, potentially malicious, mobile code. Because all of the popular web browsers execute code from multiple sources together in the same JVM, many of the security guidelines for mobile code are focused on preventing manipulation of your objects' state and behavior by adversaries who have access to the same virtual machine where your program is running.