[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: JavaHL: Bindings for svn_client_diff_summarize()

From: Daniel Rall <dlr_at_collab.net>
Date: 2006-12-08 01:32:02 CET

On Thu, 07 Dec 2006, Daniel Rall wrote:

> On Thu, 07 Dec 2006, Mark Phippard wrote:
>
> > On 12/7/06, Mark Phippard <markphip@gmail.com> wrote:
> > >
> > >On 12/6/06, Daniel Rall <dlr@collab.net> wrote:
> > >
> > >> On Wed, 22 Nov 2006, Mark Phippard wrote:
> > >>
> > >> > Daniel Rall wrote:
> > >> > >On Tue, 21 Nov 2006, Mark Phippard wrote:
> > >> ...
> > >> > >>Since you are doing some stuff with JavaHL lately, one to add to
> > >> your
> > >> > >>to-do list would be some kind of support for the svn diff
> > >> --summarize
> > >> > >>option. I think this is something we could use in Subclipse to
> > >> improve
> > >> > >>performance when comparing two tags of a large project.
> > >> ...
> > >> > ... it would return an array of the items that have been modified,
> > >> > or maybe it would send notifications that would have to be trapped.
> > >> > Either way, it would be something really different from what diff
> > >> > does now.
> > >> >
> > >> > Ideally, I'd probably like the API to return an array of some sort of
> > >> Java
> > >> > objects that describe the item and the type of change etc.
> > >>
> > >> Mark, I've written the code for the diff summarize JavaHL bindings:
> > >>
> > >> http://svn.collab.net/repos/svn/branches/javahl-diff-summarize
> > >>
> > >> When you have a moment, I'd appreciate review of at least the Java
> > >> APIs. The JUnit test case in BasicTests.java shows how to use it, and
> > >> return all diff summaries in a list.
> > >>
> > >> Unfortunately, there's a bug hiding somewhere in my implementation
> > >> which I haven't been able to shake out. The bug is triggering an
> > >> exception (in native code, I think), which subsequently causes a JVM
> > >> segfault when JavaHL's error handling code calls the
> > >> env->DescribeException() JNI routine to output a stack trace for the
> > >> exception. I've spent quite a bit of time on this bug, and could
> > >> really use some extra eyes.

I found and fixed this bug (thanks to glasser for helping me narrow
things down), and merged the branch into trunk in r22601.

Mark, could you take a look at the DiffSummaryReceiver and DiffSummary
APIs, and make sure they're in usable shape? I'm especially
interested if you think we should pull the DiffSummary.DiffKind class
out and make it more like the other classes which correspond to C
enums which are found in JavaHL.

  • application/pgp-signature attachment: stored
Received on Fri Dec 8 01:35:51 2006

This is an archived mail posted to the Subversion Dev mailing list.