Quoting Karl Fogel <kfogel@red-bean.com>:
> I wrote:
> > I brought that patch up-to-date for trunk tonight -- I'm still working
> > on it (only 4 failing tests now, down from 16!), but I like to keep
> > you informed of progress. The new patch is at http://pastebin.ca/420554.
>
> Lieven, I have to head to bed now. Here's an even more recent version
> of the patch (so ignore the above pastebin URL). It's still failing
> the same four tests, but I fixed some other things. If you have a
> chance to work on this while I'm asleep, great. If not, no worries,
> I'll work on it some more.
We're both working on the same thing then (most of your patch was already fixed
in my wc). So I'll let you do the API upgrade work and base my changes to
'status --depth' on your API changes.
I can only continue to work on it tonight (in 7 hours), so if you don't sleep
too long you won't miss anything ;)
> Note that the behavior currently implemented in the patch is the
> "dumb" -N behavior (that is, --depth=files). In fact, that's why two
> stat_tests are failing (24 and 26). Once we get the patch to match
> your proposed behavior, I expect those two to pass. (I haven't taken a
> look at the two depth_tests failures yet.)
I have those changes in my wc.
> + /* The overall depth of this edit (a dir baton may override this).
> + *
> + * If this is svn_depth_unknown, the depths found in the working
> + * copy will govern the edit; or if the edit depth indicates a
> + * descent deeper than the found depths are capable of, the found
> + * depths also govern, of course (there's no point descending into
> + * something that's not there).
> + */
> + svn_depth_t default_depth;
> +
> /* Do we want all statuses (instead of just the interesting ones) ? */
> svn_boolean_t get_all;
>
> @@ -111,6 +119,14 @@
> directory. */
> struct dir_baton *parent_baton;
>
> + /* The ambient requested depth below this point in the edit. This
> + can differ from the parent baton's depth (with the edit baton
> + considered the ultimate parent baton). For example, if the
> + parent baton has svn_depth_immediates, then here we should have
> + svn_depth_empty, because there would be no further recursion, not
> + even to file children. */
> + svn_depth_t depth;
It's not clear to me why you need both a depth and a default_depth field. I
probably didn't dive into the status code deep enough yet to understand, but I
have all the status tests working with the extra default_depth field.
Lieven
----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Apr 2 12:11:44 2007