>
> Yes, I did. It does not address the limitations of which I am speaking.
If ffap-support was extended to search in the same directories as JDE
uses
it would be able to find any java filename.
I will extend it with that functionality tomorrow.
The second problem as you mention in another posting is the possibility
for multiple
matches - this I do not have a solution for. But I'll look into it :)
> >And does JDE examine the JAR files for .java files and extract them to
> >be shown ?
> >
>
> No, it does not extract source files from jar files. I am referring to the
> fundamental flaw in any solution that depends on the classpath to find
> source files, namely that the classpath specifies the locations of CLASSES
> not SOURCE files on a user's system. It is not generally the case that
> classes and their source files reside in the same location.
Well - as said before I used CLASSPATH as it was a quick way to solve
Ivan's problems :)
>
> After Beanshell startup, the time to resolve the class name is imperceptible.
>
> >+ Does not require the java file to be compiled
>
> How often do you need to find a file that has not been compiled?
Well - not often, but I had a few of them today when I was refactoring a
class into multiple classes - and it was
quite usefull to be able to just press C-x f on the class name and let
ffap-java-support find them.
> >+ Beanshell does not (yet?) support automatically loading class'es
> > which did not exist when it was invoked
>
> The command jde-wiz-update-class-list loads classes that are compiled after
> the Beanshell is invoked. It would be trivial to invoke this automatically
> whenever a Java class is compiled.
Ok - sorry, it did not know that.
> >
>
> The only advantage I can see is that it does not require the Beanshell. I
> personally cannot live without the Beanshell so the advantage is no
> advantage to me. I have no objection to your posting your ffap-based
> solution to the JDE mailing list as long as you are very clear about its
> actual limitations and do not make incorrect or unfairly disparaging
> statements about the capabilities of alternative solutions.
Ok - I will make an extra effort for this in the future.
> I also do not object in principal to offering an ffap-based solution as
> part of the JDE. However, I will do so only if it offers at least the same
> capabilities as the existing jde-show-class-source-command with the same
> ease of use, that is, it must be capable of finding the source for any
> unqualified class name at point that is listed in
> jde-db-source-directories, which by the way is a list of package directories.
That will be possible tomorrow with just a few lines of extra elisp :)
With hope
Max
