you are here: projects     select style: classic proton

json comment stripper

json comment stripper

A minimal JSON comment stripper written in Groovy. Eats up all comments a given JSON file brings along. Won't change line or column numbers.

The archive also contains a small GraphML file visualizing the underlying state machine.

fast markov algorithm interpreter (fmain)

fmain at work

If you don't know what this program could do it will be quite useless for you. Although one could use it to parse and modify all kinds of files there are a lot of other tools which normally fit much better. Let's just say that markov rules don't support something like regular expressions.

As you can see in the screen shot the program doesn't come along with a gui. You simply have to hammer the rules in a file using your favourite editor. The rules file and also the input and output files are specified via command line parameters before running the program. Feel free to use "fmain --help" to get some information how this could be achieved.

educational markov algorithm interpreter (emain)

emain at work

While the fmain described above is optimized on speed this tool is meant to be used in education and for debugging markov rules. For these puposes the emain got an easy to use gui where the user is able to input character data and rules in visible text fields. Furthermore there are buttons and switches to start, pause or control the appliance of rules in many ways. You may want to browse the directory listing to have a look at some sample text files.

The program isn't compiled for a special hardware or system. All it needs to run emain is a modern web browser capable of latest html and some javascript. Any version of Mozilla should do fine. It also works with any newer Internet Exploder or Konqueror.

post calculus interpreter (pocain)

pocain at work

Similar to the emain above this litte tool comes along with an easy to use interface. As anybody can guess from the name it works as an interpreter for post calculus rules. The user is able to switch between two different output modes in combination with a special step mode as a further help in your crazy debugging sessions. You may want to browse the directory listing to have a look at some sample text files.

The program isn't compiled for a special hardware or system. Any modern web browser can be used as a runtime environment. Firefox and Konqueror have been tested and work just fine. You should avoid Internet Exploder 6 or below when possible for it has problems rendering the output and the step mode behavior seems frelled too.

puzzle solver

puzzle solver

The screen shot of this program might not look too spectacular. Nevertheless believe me, the whole thing was more difficult to code as most people - including myself - think at first thought. Remaining people will say: Yes, of course it can't be that simple, as you have to use three level backtracking to kill that beast.

You can find some more lines speaking about idea and functionality of the puzzle solver by reading in the gallery section. Just like the fmain on top of the page this program was also written in Pascal. To build the binary you will need the gpc and also an ncurses-devel package installed on your system.

dust map editor

dust mapedit

This web browser based application was created by me supporting a friend's multi player browser game project called Dust. Although the project isn't online anymore and the map editor is somewhat incomplete it is still a good example for what one can do using a few lines of a good script lanquage like JavaScript and a standards compliant browser like Mozilla.

No further explanations needed. Simly click the link following this text to test it out. Don't forget to activate JavaScript in your Browser and don't try this using Internet Exploder. Standard compliance, remember?

cheap assembler project

cheap simul.

What we have here is a three in one package containing an assembler, a simulator and an extensive documentation to both written mostly in german lanquage. While a friend focused on the assembler part I did the remaining work. The picture on the left shows the mozilla based simulator in action.

First we defined an easy to understand processor comparable to the Intel 8080. This means we wrote a list of commands it should support and such things. After that the real work started. My friend wrote the whole lex and yacc code while I developed the simulator we wanted to let the assembler created machine code run on. This way it was possible to test wether the assembler produced valid code. Furthermore it would have been simply a lot less fun to inspect the assembler output using a boring hex viewer.