jde-xref-update will only work if you edit all your files in the same
instance of emacs, then update.  I would advise doing a full rebuild
nightly while your machine and emacs is idle (use the emacs timer
functions).

If the creation of the database is slow, make sure your
jde-parse-class.el is byte-compiled into jde-parse-class.elc.  I've
also noticed that it works faster on a brand-new emacs then one that
already has a bunch of stuff in memory.

aaron armstrong <[EMAIL PROTECTED]> writes:

> Andrew,
>
> Thanks - it must be that my xref-db is out of date.  Does
> jde-xref-update update the db or do I need to rebuild it?  I a working
> on a very large code base so is there any way to just update what has
> changed or just update files/packages i specify?
>
> thanks
> aaron
>
> At Thu, 19 Feb 2004 14:42:09 -0800,
> Andrew Hyatt wrote:
>> 
>> 
>> From where are you doing a C-c C-v a?  From inside a function should
>> work.  Also, to get useful results, you need classfiles that match up
>> with your sources.  So if your classfiles are out of date by a while,
>> you may see some strange things.
>> 
>> I haven't heard of problems like you've described, however.  When you
>> see the call tree, you should get children grouped by class, where
>> the children are the functions that call the function you are
>> interested in.  Clicking on the children should take you right to the
>> caller.
>> 
>> I'm not sure if any of this will help.  If you think the generated
>> xref db must be wrong, then I can tell you how to manually go in and
>> verify that it is wrong.
>> 
>> aaron armstrong <[EMAIL PROTECTED]> writes:
>> 
>> > Andrew,
>> >
>> > Thanks for helping me create the xref db.  Now I'm trying to use it
>> > and I'm a little confused by the way it works.  When I display the
>> > call tree I get information about classes that don't actually call the
>> > method I'm interested in.  Then if I do C-c C-v a (xref-first) it
>> > takes me to the beginning of the class rather than the place that
>> > actually calls the method.  Is this the expected behavior?
>> >
>> > thanks
>> > aaron
>> 
>> 

Reply via email to