[Rprotobuf-yada] [patch] More graceful error handling, support import patterns other than .proto$, and add simple repeated field unit test

[Murray Stokely]

I added one more unit test that I had neglected to svn add in that
previous patch.  I've updated it at the same location with another
tiny test to make sure that an error is returned rather than a
segfault when there is a problem importing proto files :

    http://www.freebsdmall.com/~murray/patch_by_Murray_rprotobuf_r406.diff

Since you've already applied it, probably just the new standalone file
would be easier :

    http://www.freebsdmall.com/~murray/runit.import.R
(for pkg/inst/unitTests)

I have nothing else outstanding so sure, rolling up a release would be
great.  I will probably use RProtoBuf more heavily this week than I
have in the past so I may run into other minor things, but nothing
known at the moment.

             - Murray


On Mon, Jan 10, 2011 at 6:29 PM, Dirk Eddelbuettel <edd at debian.org> wrote:
>
> Hi Murray,
>
> On 10 January 2011 at 18:04, Murray Stokely wrote:
> | Hi Dirk, Romain, et al.
> |
> | Turns out I'm able to use your modified RSourceTree just fine so I
> | won't need to make any significant changes to the way you import the
> | protocol buffer definitions, but I do have a small patch to make
> | things a little easier.
> |
> |    http://www.freebsdmall.com/~murray/patch_by_Murray_rprotobuf_r406.diff
> |
> | While here, I added a test for the simple issue we used to have with
> | RProtoBuf v0.1 but which you subsequently fixed.
>
> Really nice simple patch -- nice idea with the pattern argument!  Patch
> applies cleanly like a charm and you even adopted my somewhat random doubly
> redundant (patch, rprotobuf) naming pattern in the patch/ directory ...
>
> Do you think that's it for now?  In which case I'll probably roll up a
> release with your two patches.
>
> Thanks again, Dirk
>
> --
> Dirk Eddelbuettel | edd at debian.org | http://dirk.eddelbuettel.com
>