Free java class diagram software for linux

Class::Inner is a perlish implementation of Java like inner classes.

SYNOPSIS

use Class::Inner;

my $object = Class::Inner->new(
parent => ParentClass,
methods => { method => sub { ... } }, },
constructor => new,
args => [@constructor_args],
);

Yet another implementation of an anonymous class with per object overrideable methods, but with the added attraction of sort of working dispatch to the parent classs method.

METHODS

new HASH

Takes a hash like argument list with the following keys.

parent

The name of the parent class. Note that you can only get single inheritance with this or SUPER wont work.

methods

A hash, keys are method names, values are CODEREFs.

constructor

The name of the constructor method. Defaults to new.

args

An anonymous array of arguments to pass to the constructor. Defaults to an empty list.

Returns an object in an anonymous class which inherits from the parent class. This anonymous class has a couple of extra methods:

SUPER

If you were to pass something like

$obj = Class::Inner->new(
parent => Parent,
methods => { method => sub { ...; $self->SUPER::method(@_) } },
);

then $self-gtSUPER::method almost certainly wouldnt do what you expect, so we provide the SUPER method which dispatches to the parent implementation of the current method. There seems to be no good way of getting the full SUPER:: functionality, but Im working on it.

DESTROY

Because Class::Inner works by creating a whole new class name for your object, it could potentially leak memory if you create a lot of them. So we add a DESTROY method that removes the class from the symbol table once its finished with.

If you need to override a parents DESTROY method, adding a call to Class::Inner::clean_symbol_table(ref $self) to it. Do it at the end of the method or your other method calls wont work.

clean_symbol_table

The helper subroutine that DESTROY uses to remove the class from the symbol table.

new_classname

Returns a name for the next anonymous class.

The Java Binary Enhancement Tool (JBET) is a general Java program analysis and manipulation tool. Existing class files can be disassembled, reassembled, or edited programmatically through the JBET API. JBET can also be used to create new Java class files from scratch. JBET uses a convenient internal representation of all the contents of Java binary (.class) files, allowing the user to edit the classes easily, in a structured manner.

JBET was developed as part of the DARPA Self-Protecting Mobile Agents project under the OASIS and Active Networks programs (contract number N66001-00-C-8602) in order to study automated software obfuscation.

The Java language was chosen for this project because of the (relative) ease of constructing binary editing tools provided by the large amount of type information present in the class files. Our two reports, the Obfuscation Techniques Evaluation Report, and the Obfuscation Report, are available from the download area. The obfuscation tool developed is not part of this release.

JBET was also used in the DARPA/AFRL Survivable Server project (contract number F30602-00-C-0183) to add additional security checks to the Java Standard Library. (The Java SecurityManager API does not support many desirable security checks, such as continued authorization of file accesses after opening.)

JBET was used to replace the native method references in the Java standard library with stubs that call a pluggable security policy. This tool, called Jpolicy, is also available for download at this website. Jpolicy is very incomplete at this time, but may be interesting to those working in Java security or changing the standard library themselves.

The internal representation of Java class files used by JBET is intented to make it easy for programmers to write Java binary code transforms. Each element of Java class files has a corresponding internal data structure: ClassInfo for entire classes, MethodInfo for methods, FieldInfo for fields, Snippit for code blocks, and Instruction for individual instructions. Snippit and Instruction understand Java opcode syntax and semantics, allowing automated creation of valid Java programs. A Java-compatible class verifier is also included.

Some code transforms are difficult to program directly by manipulating Java instructions. For those transforms, a directed acyclic graph (DAG) representation of code is available. In the DAG representation, each basic block has a corresponding DAG, with a set of input and output nodes. Edges in the graph connect "producer" nodes (such as constants, or the result of calculations) to "user" nodes (such as method calls or other calculations). Methods are divided into basic blocks and control flow is stored at the basic block level (possible because Java has only fixed jump targets)

JBET requires a Java 1.4 virtual machine to run, although it can operate on class files from earlier Java versions. The packaging and build environment supplied supports Linux and Windows with Cygwin; however, the build process is simple and could be performed manually on other platforms. Perl is required for regression testing.

Jpolicy requires a Java 1.4 virtual machine to build, either Linux or Windows NT/XP with Cygwin. gcc is required for building on Windows (supplied with Cygwin). The runtime system can be either Java 1.3 or 1.4 (with Suns JVM only), running on Linux or Windows NT/XP. Windows 9x and Windows 2000 may work as well, but have not been tested.

Installation

1. Install jdk 1.4.1.
2. Set CLASSPATH to jdk1.4.1/jre/lib/rt.jar
3. cd src; make
4. If that didnt work, examine the makefile. java or javac may not be in the path.
5. To build a jar file that can be used with "java -jar jbet.jar", run "make jar".
6. If you have perl installed, run the tests with "make test".
Optionally, run "make regen; make test".

Make a symbolic link from jbet3/bin/jbet to somewhere in your path.

Usage

JBET uses the JNI format for class names, and JNI type and method descriptors. For a summary of this syntax, use jbet help syntax. Suns JVM specification may also be helpful.

To look at a class disassembly, use jbet print. Try disassembling a class you have source for, and was built with debug info (-g): jbet -P < classpath > print < classname >. Suns JVM specification has an instruction reference.

Clara is a help-documentation tool for C++ or java developers. Basically she paints a diagram - a clara diagram - from a class.

With clara you can visualize your class as if it were a chip with pins for input - representing the input parameters of the methods - and pins for output - representing the return type of the functions-.

Lets see for example the diagram that clara would draw for the following C++ sample class

class sampleInterface
{
public:

void init (MyStructure * pstruct);

void setPoint (float x, float y);

float getX ();
float getY ();

void processIt ();
};

What are these pins for ? Are we going to connect them somehow ? Not at all, it is just a convenient way to represent what the class offers in a very compact form.

Of course the diagram will be drawn by clara automatically, there is no need (and no chance) of manually drawing anything.

Clara gets the needed information in two ways : using reflection in java and parsing the header file in C++. Clara can also draw a call-diagram of the class if the sources are found, note though that for the call-diagram an extra tool is needed (dot.exe from http://www.graphviz.org) and it has to be installed apart.

Class::Declare::Attributes is a Perl module with Class::Declare method types using Perl attributes.

SYNOPSIS

package My::Class;

use 5.006;
use strict;
use warnings;

use base qw( Class::Declare::Attributes );

# declare the class/instance attributes
__PACKAGE__->declare( ... );

#
# declare class/static/restricted/etc methods of this package
#

sub my_abstract : abstract { ... }
sub my_class : class { ... }
sub my_static : static { ... }
sub my_restricted : restricted { ... }
sub my_public : public { ... }
sub my_private : private { ... }
sub my_protected : protected { ... }

Class::Declare::Attributes extends Class::Declare by adding support for Perl attributes for specifying class method types. This extension was inspired by Damian Conways Attribute::Handlers module, and Tatsuhiko Miyagawas Attribute::Protected module. The original implementation used Attribute::Handlers, but now simply refers to attributes.

The addition of Perl attribute support (not to be confused with object attributes, which are entirely different, and also supported by Class::Declare) greatly simplifies the specification of Class::Declare-derived class and instance methods. This should aid in the porting of existing code (Perl, Java and C++) to a Class::Declare framework, as well as simplify the development of new modules.

With the addition of Perl attributes, Class::Declare methods can now be written as

sub method : public
{
my $self = shift;
...
}
instead of
sub method
{
my $self = __PACKAGE__->public( shift );
...
}

Java::Import::Examples is an example of how to use Java::Import to call into Java classes.

Making RMI calls from Perl

One nice thing about Java is the extremely straight forward manner in which it allows you to make calls to remote objects residing on distant servers. Many people use this ability as a point of integration between their system and a posible third party or legacy system. Unfortunatly, for the most part the ability to make calls to these remote objects is something that only other Java applications can do without making a big fuss about it. However, we can now do this from Perl.

Assume that I have a RMI server from which I can get data from in the form of simple Java Beans. One method of doing this is to use the GCJ::Cni library and natively compile and wrap a set of Java Classes which can then be used from my Perl script. However, there is an easier way.

Lets begin by descriping the Java interfaces that well be working with. We first have the Remote Interface that we will be interacting with:

import java.rmi.*;
public interface RemoteInterface extends Remote {
public SomeBean getMessage(String seedMessage) throws RemoteException;
}
And we also have the bean we will be asking for:
import java.io.*;
public class SomeBean implements Serializable {
private String value;
public SomeBean() {}
public void setValue ( String _value ) { ... }
public String getValue ( ) { ... }
}
As far as any Perl client program is concerned this is all we need to know about.
Now all there is left to do is write a client:
use Java::Import qw(
java.rmi.Naming
);

my $remote_interface = java::rmi::Naming->lookup(jstring("//localhost/Home"));
my $bean = $remote_interface->getMessage(jstring("Hi there"));
print $bean->getValue(), "n";

Thats all there is. Notice that all we had to tell Java::Import about was java.rmi.Naming, this is because it was the only class we used by name in our Perl code, every other Java class (the objects held by $bean and $remote_interface) was returned by some other method call originating from java.rmi.Naming.

In order to run this example we have to make sure everything is in its place. We start by compiling the client code. Assuming that we have been given a Stub class file (RemoteObject_Stub.class) we can do the following:

gcj -C SomeBean.java RemoteInterface.java
fastjar -cvf client.jar SomeBean.class RemoteInterface.class RemoteObject_Stub.class

Making sure our server is reachable we can then run the client code:

CLASSPATH=client.jar perl client.pl

Thats it, you should then see the message returned by your server.

Java::Import is Perl module to use Java classes in Perl.

SYNOPSIS

use Java::Import qw(
some.package.SomeClass
);

my $instance = new some.package.SomeClass();
$instance->someMethod();

my $ret_val = some::package::SomeClass::someStaticMethod();
$ret_val->someMethod();

$ret_val2 = $instance->someOtherMethod($ret_val);
$ret_val2->someMethod();

my $java_array_ref $instance->someMethod2();
foreach my $obj ( @$java_array_ref ) {
$obj->someMethod();
}

The purpose of this module is to provide a simple method for using Java classes from a Perl program while using the latest in Open Source Java Technology. Thus, this module makes great use of the GNU Compiler Tools for Java in its implimentation.