On Thu, Jul 3, 2008 at 9:57 AM, Alberto Gimeno <gimenete_at_gmail.com> wrote:
> The cache is calculated just once. But that first time it has to do
> several calculations. For example when I get this log messege:
>
> D /foo/bar
>
> In my implementation I have to update all the subfiles and subfolders
> of /foo/bar to mark them as deleted in that revision. However the
> second time, the cache just needs to query for updates. For example in
> a repository with 4000 revisions the first time it needs to cache
> those 4000 revisions. But the next time the cache just will need to
> cache 10, 20, 30 revisions. However, yes, for the first time it takes
> too much time.
>
> Now I'm using PreparedStatements and I reuse them. Maybe I could use
> batch operations. But I'm not sure.
>
> I want to take some metrics on different repositories. But I haven't
> had time these days.
>
> Also I would like to update the cache in chunks. For example from 100
> to 100 revisions. By doing this I could update the monitor progress
> bar between those chunks. Now I'm getting all the log messeges inside
> one array by using
>
> client.getLogMessages(info.getRepository(), latest, SVNRevision.HEAD);
>
> Where "latest" is the latest revision number I have in the cache. Is
> there any way to know which revision number has the HEAD revision?
Accurate and detailed progress information certainly makes any
operation feel faster. Do you need to know HEAD before you call the
API? I'd think you could just look at the array you get back and see
how many entries it contains or even peek at the last entry. If you
need an idea of what HEAD is before you run, you could call the info
API and pass it HEAD and then see what the revision is that you get
back.
--
Thanks
Mark Phippard
http://markphip.blogspot.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subclipse.tigris.org
For additional commands, e-mail: dev-help_at_subclipse.tigris.org
Received on 2008-07-03 16:05:56 CEST