Re: svn commit: r10008 - trunk/subversion/libsvn_repos

["Peter N. Lundblad" <pe...@famlundblad.se>]

On Wed, 16 Jun 2004 lundblad@tigris.org wrote:

> Author: lundblad
> Date: Wed Jun 16 01:19:44 2004
> New Revision: 10008
>
> Modified:
>    trunk/subversion/libsvn_repos/rev_hunt.c
> Log:
> Make svn_repos_trace_node_locations work for future versions even if the path
> isn't in the for the filesystem presents it.
>
> * libsvn_repos/rev_hunt.c (check_path_ancestry): Get the filesystems form
>   of the path during the first history iteration.
>
>
> Modified: trunk/subversion/libsvn_repos/rev_hunt.c
> Url: http://svn.collab.net/viewcvs/svn/trunk/subversion/libsvn_repos/rev_hunt.c?view=diff&rev=10008&p1=trunk/subversion/libsvn_repos/rev_hunt.c&r1=10007&p2=trunk/subversion/libsvn_repos/rev_hunt.c&r2=10008
> ==============================================================================
> --- trunk/subversion/libsvn_repos/rev_hunt.c	(original)
> +++ trunk/subversion/libsvn_repos/rev_hunt.c	Wed Jun 16 01:19:44 2004
> @@ -288,6 +288,13 @@
>
>    SVN_ERR (svn_fs_node_history (&history, root, fs_path, lastpool));
>
> +  /* Since paths that are different according to strcmp may still be
> +     equivalent (due to number of consecutive slashes and the fact that
> +     "" is the same as "/"), we get the "canonical" path in the first
> +     iteration below so that the comparison after the loop will work
> +     correctly. */
> +  fs_path = NULL;
> +
[...]

I choose this solution to the problem I posted about yesterday. It makes
test pass on ra_local again (with the yet uncommited rest of the
get_locations patch, no break on trunk:-).

Now, there is just one fail on ra_dav. Thanks to everyone who writes
tests! They are just terribly useful!

Regards,
//Peter

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org