Project Contours is now on track

Yes, there has been silence for some time. Cause of that are the upcoming exams this weekend in lectures which awesomeness-level lies below that of a rotting potato.

Nevertheless, there has been some progress. I finally decided a name and hoster for my project which I've been pondering now for almost two semesters at University. And with some traces of proudness I finally present you the

Contours Framework

To admit, the hardest part was really to find a proper name for it, which should hint a bit into the direction of the topic without colliding with other projects. Another earlier try was "Shapes" but there are already several projects around with that term in their name and which are probably more popular on search terms than my project will ever be.

In more details, the project is about creating a framework which allows Java application developers to create a file-reader for arbitrary binary files which then allows easy access to the content data from the file. This is done by providing a custom file-description language in which the structure of the file can be laid out. Then this format description is taken by the Contours framework which produces all necessary Java code for a file reader for files of this structure. This reader can then be easily be integrated into the own Java application.

I know, it's not a very common problem this addresses and there are other preferable ways for most similar problems. But I needed something which makes it easy to follow changes in file formats without the need to regenerate and review large parts of existing Java code. Furthermore when I'm dealing with more than one or two different formats, the maintenance of several different readers written by different people gets unhandy quite fast. And finally one of the goals was to produce code which allows high processing speed. When I started my project, there was no framework around addressing all of these points (and there still is none). Nowadays there are some available (one of them being Preon) but none of them seems to fit the performance requirement. And I've gotten already quite far when I discovered those, so I chose to continue with it and if it's just for learning purposes.

Nevertheless, the page is up, some content already present (notably the architecture description of Contours) and all source committed to the Contours SVN source repository. Feel free to have a look at it and if you have any comments, comment here or raise an Issue (when I come around to enable and configure that part of the project properly).


Similar entries