[XQZone General] MLJEdit XQuery plugin

[Margaret Leber]

Ron Hitchens wrote:

>    I wrote both the jEdit plugin and XQRunner.  That sort
> of exception generally indicates a classpath problem, but
> I'll look into it.
> 
>    Which OS are you running?  Which version of the XDBC
> jars are you using?  Can you double check that both xdbc.jar
> and xdmp.jar are in your classpath as well?

xdbc.jar and xdmp.jar are both the versions that shipped with the 
plug-in--dated 9/23/04--, and I copied them to the .jedit\jars directory 
along with the plugin, and also tried moving them to Program 
Files\JEdit\jars...same result. I'm currently testing on Windows XP.

It sure would be a good thing from a debugging point of view if, when 
XQRunner catches an Exception and throws a new Exception, that the new 
Exception include all the information from the original Exception.... I 
scratched my head for a while over the stark message "not found", which 
*looks* like it's coming from e.getMessage(). Calling getMessage() on 
ClassNotFoundException after Class.forName() returns the class name that 
wasn't found...so I'm thinking we may be catching something else--maybe 
a nested XQException? Having the xqrunner.jar at 15k  vs. xqrunner at 
60k in the 0.7 version suggests to me that things have changed *a lot* 
in xqrunner since last September, so looking at the 0.7 source may be 
leading me astray.

Using the exception chaining mechanism introduced in Java 1.4 might be 
helpful too, but I don't know what your legacy Java platform support 
goals might be. At the very least e.getClass().getName() should be in 
the message of the new Exception.

To really shoot this issue properly I'm going to have to get to where I 
can compile xqrunner and still reproduce the error...so maybe I should 
recompile the plugin too.

--
  Margaret Leber CCP, SCJP, SCWCD
  610-630-9862

hire me! http://voicenet.com/~maggie/mslresume.pdf