webloupe / News: Recent posts

It includes a new multi-threaded crawler, which has been programmed by myself. It's based on new concurrent programming features of the JDK 5.0.
The TouchGraph classes for visualization of information architecture have also been integrated well, extended with the possibility to make a screenshot of the TouchGraph and save it as PNG file.
Further visualization, besides our standard JTree, is possible by generating a GraphXML file with WebLoupe, for use with the open-source HyperGraph applet, which comes as part of the WebLoupe packages.
Furthermore, a table rich of info about Web site reachability and content is shown after the crawler has stopped.
You can launch the browser with an URL by doubleclicking on a JTree node.
The application reads and writes property files now, so it will always remember your last settings.... read more

The project team has released version 0.1 of WebLoupe, the java-based open source web crawler and information structure visualization tool.

This release reflects a very early state of development. The crawling is not multi-threaded yet. Therefore, the crawling performance is not sufficient. In a future release of WebLoupe, it's planned to incorporate a multi-threaded crawling algorithm, by implementing code of an open source crawler.... read more

Because of performance issues when crawling websites, the project team is going to change the web spider algorithm of webloupe. They're looking for an open source java web spider which supports multithreading and is suitable for integration.
For visualization of the information structure, they're trying to integrate Touchgraph (http://touchgraph.sourceforge.net/).
In the next few days, after assembling the sources and implementing the core functionalities, they will commit a new version of webloupe.... read more