And with this, I did a simple commit to the server!
Just tried one with an added directory and files and a change, and it
broke... sigh. However, it should be minor, as I've previously tested the
make-dir and add-files before; it is probably a matter of some small tweaky
issue (looks like a version URL parsing issue on the root resource).
Once I fix that last bit, then commit should be "done enough", with the
caveat for lurking bugs. (the part that bothers me is that server isn't very
DAV-compatible with its response generation, but we don't actually need that
right now)
After this comes the update. This "should" be straight forward since I'm
(effectively) going to be tunneling a custom request/response through a
WebDAV REPORT method. That custom stuff is exactly what Ben has already
done, but with network marshalling between ra_dav and the server (as opposed
to Ben's function calls between ra_local and fs).
I'm hoping to do the update before I leave for NYC on Thursday. I think that
will put us at M2. (oh; need to look at "import")
Cheers,
-g
On Mon, Apr 16, 2001 at 01:00:04PM -0000, gstein@tigris.org wrote:
> User: gstein
> Date: 01/04/16 06:00:04
>
> Modified: subversion/mod_dav_svn dav_svn.h repos.c
> Log:
> Check the Content-Type header to see what is being PUT to the server, and
> process the (diff) stream accordingly.
>
> * dav_svn.h (struct dav_resource_private): add (hacky) is_svndiff flag.
>
> * repos.c (struct dav_stream): add a new writable stream field.
> (dav_svn_get_resource): check the Content-Type header and set the
> is_svndiff flag.
> (dav_svn_open_stream): open an svndiff->txdelta stream if is_svndiff has
> been set.
> (dav_svn_close_stream): close the wstream, if present
> (dav_svn_write_stream): write to the svndiff/txdelta stream, or create a
> window for writing plain text.
>...
--
Greg Stein, http://www.lyra.org/
Received on Sat Oct 21 14:36:28 2006