>>> David PONCE <david.ponce@...> seems to think that:
>> I've introduced a new kind of data cache behind the API fcn in
>> semantic-fw.el. The starting point is:
>> `semantic-cache-data-to-buffer'. This will link a data structure to
>> a specific region of a buffer. You can retrieve the data with
>> `semantic-get-cache-data' iff the point is in the region.
[ ... ]
>> This is very useful with the idle scheduler. Both summary mode and
>> the completion mode need that context. This basically cuts the
>> calculation time spent in the idle scheduler in half. Huzzah!
>That looks like a very good idea!
>I have a suggestion: to replace `tag' in that data cache API with
>cache name, or something like that, to avoid confusion with the
>general meaning of `tag' in semantic.
>Also it looks that there is a small bug in the
>`semantic-cache-data-post-command-hook' loop, which only look at the
>first overlay in `semantic-cache-data-overlays'.
>Finally, in `semantic-cache-data-to-buffer' I think it would be
>better to check if LIFESPAN is valid before to create the overlay and
>push it in `semantic-cache-data-overlays'.
[ ... ]
Thanks for checking up on it. You are right makes sense not to use
tag. We use 'tag' for this naming purpose in our object system here
at The MathWorks.
You other identifications are correct as well. I do appreciate your
Eric Ludlam: zappo@..., eric@...
Home: http://www.ludlam.net Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net GNU: http://www.gnu.org