Hi Stan,
That's definitely usable. We'd already made something like that ourselves for the pIMPed Meta-Environment, but yours is better.
Cheers,
Jurgen
On Fri, Oct 10, 2008 at 1:00 AM, Stan Sutton
<suttons@xxxxxxxxxx> wrote:
Hi All,
In response to a suggestion (and contribution)
in Eclipse bugzilla bug #245296, I've created a subtype of org.eclipse.imp.builder.MarkerCreator
called MarkerCreatorWithBatching. As you might guess, this marker
creator batches up error messages without creating markers immediately,
then creates the markers all in one workspace operation when a flush()
method is called. It also supports the setting of limits on the number
of errors of various severities that are to be processed.
In our usual example service implementations,
we only use MarkerCreator in the builder and poor man's compiler. I've
modified the templates for both of those to include declarations of a marker
creator of each type, with the declaration of the newer type commented
out, so that it's easy to experiment with them.
I did play around a little, trying both
types of marker creators on a LEG file with about 200 errors. I couldn't
see any difference in performance, but I didn't time it precisely, and
maybe that's not the most stressful case (the original bugzilla report
cited files with many hundreds of errors). Anyway, now we have a
couple of very different implementations to choose between.
Check it out, literally and figuratively,
if you're interested, and let me know about problems, suggestions, etc.
Thanks,
Stan
Stan Sutton, Ph. D.
IBM T. J. Watson Research Center
19 Skyline Drive, Hawthorne, NY 10532 USA
telephone: 1-914-784-7316, FAX: 1-914-784-7455, T/L 863
e-mail: suttons@xxxxxxxxxx, Stan Sutton/Watson/IBM@IBMUS
_______________________________________________
imp-dev mailing list
imp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/imp-dev
