On 12 November 2011 03:37, Bill van Melle <[email protected]> wrote: > In another direction, I suppose one could write a Java program linked > with the Pivot libraries and using reflection to generate the same > sort of information in real time. I'm sure there already exist > browsers of Java classes, maybe even ones oriented to showing the > "bean" view of classes, but the advantage to a dedicated one is that > it could also incorporate skin information into a unified > presentation.
This is the approach I have taken, but have been too busy to even get the code on the mailing list for discussion. A few tweaks to BeanAdapter opens things up nicely. The functionality of its private inner PropertyIterator class can be exposed, after a modification so that it works on a java.lang.Class directly, rather than an instance of a 'bean'. http://svn.apache.org/repos/asf/pivot/trunk/core/src/org/apache/pivot/beans/BeanAdapter.java Similarly, code already exists to enumerate the events. http://svn.apache.org/repos/asf/pivot/trunk/tutorials/src/org/apache/pivot/tutorials/explorer/tools/EventLogger.java Ultimately, I would like to see all of the BXML/bean related data available from a simple call to a single method provided as part of Pivot. This would be an authoritative source of that metadata which could then be presented as required. My efforts initially led me to generating some Javadoc type HTML documentation with searchable & sortable tables, but then I ditched that and went with a simple and far more functional Pivot app. After all, if I need to know Pivot bean info, it is most likely when I am at a dev box and developing at the time. Chris
